Tactics cannot be very important for chess

Discussion of anything and everything relating to chess playing software and machines.

Moderators: bob, hgm, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
hammerklavier
Posts: 104
Joined: Mon Nov 28, 2011 9:17 am

Re: Tactics cannot be very important for chess

Post by hammerklavier » Fri Nov 01, 2019 5:23 pm

A curious parallel between man and machine. In the fourth volume of the General Chess Treaty of Roberto Grau, the author advises the deductive method instead of the analytical one. When the machine can deduce in a way close to human there will be a quantum advance.

Chapter I

The transformation of threats.

The real difficulty of chess is in knowing the degree of transcendence that for the full gravitation of the games have the positions to be created. This makes the most inconsequential plays in appearance may have decisive importance in the course of the fights, and gives the player accentuated interest, which arises from the difficulty that these technical subtleties present, even for the most experienced player.
We have said on several occasions that chess was not that monument of complexity that some vain players suggest, and we insist on the same. Playing chess is relatively easy, and playing discreetly is not a task of giants either. It only takes a certain dedication and a lot of practice. Now that this does not mean, much less, that the difficulties are running out when the chess player advances. On the contrary, these are accentuated as it becomes subtle, and this is the paradoxical fact that the one who has the most difficulty playing, as he aspires to do, is the first-force player, as he often pretends impossible: he wants to exhaust the analysis and try to reach definitive conclusions by means of a thorough examination of future possibilities and by studying the possible moves and answers, without occasionally, making an examination of conscience and considering if indeed your brain is in conditions of Perform such an extensive task, without risks of incurring very serious analysis errors.

The Chess player must deduce without deep analysis.

This is why we have always argued that the player should, if possible, discard the analytical system to adopt the deductive. He must, in the middle game and in the opening, believe less in mathematical truth - because in reality it is very difficult to undress - and conform to the logical truth, which is always more accessible and has the virtue of encouraging deduction. It is clear that it is more dangerous, but the truth is that in this way intuition plays a more important role, and imagination finds less obstacles.
If Chess were a mathematical game, it would certainly be much more boring than in the present. The possibility that those who reason with more logic can overcome those who analyze better, or that even those who are better inspired exceed those of more logic on many occasions, make this game as delicious as inconsequential art, and I assure you perpetual youth.

Original text in Spanish...

Capitulo I

La transformación de las amenazas.

La verdadera dificultad del ajedrez esta en conocer el grado de trascendencia que para la gravitación completa de las partidas tienen la posiciones a crearse. Esto hace las jugadas mas intrascendentes en apariencia puedan tener importancia decisiva en el curso de las luchas, y brinda al jugador acentuado interés, el cual surge de la dificultad que estas sutilezas técnicas presentan, aun para el jugador mas experimentado.
Hemos dicho en varias oportunidades que el ajedrez no era ese monumento de complejidad que algunos jugadores vanidosos hacen suponer, e insistimos en lo mismo. Jugar al ajedrez es relativamente fácil, y jugar discretamente tampoco es una tarea de gigantes. Solo hace falta cierta dedicación y bastante practica. Ahora que esto no quiere decir , ni mucho menos, que las dificultades se vayan agotando cuando el ajedrecista avanza. Por el contrario, estas se acentúan a medida que se sutiliza , y resulta así el hecho paradójico de que quien tiene mas dificultad para jugar , como el aspira a hacerlo, es el jugador de primera fuerza, pues pretende a menudo imposibles: quiere agota los análisis y tratar de llegar a conclusiones definitivas por medio del examen minucioso de las posibilidades futuras y por el estudio de las jugadas y respuestas posibles, sin de vez en cuando, hacer un examen de conciencia y considerar si en verdad su cerebro esta en condiciones de realizar tarea tan extensa, sin riesgos de incurrir en errores de análisis muy graves.

El Ajedrecista debe deducir sin análisis profundos.

Es por esto que siempre hemos sostenido que el jugador debe , en lo posible, desechar el sistema analítico para adoptar el deductivo. Debe , en el medio juego y en la apertura, creer menos en la verdad matemática – porque en realidad es muy difícil desnudara- y conformarse con la verdad lógica, que siempre es mas accesible y posee la virtud de fomentar la deducción. Es claro que resulta mas peligrosa, pero lo cierto es que de esta suerte la intuición juega un papel mas importante, y la imaginación halla menos trabas.
Si el Ajedrez fuera un juego matemático , seria por cierto mucho mas aburrido que en lo presente. La posibilidad de que quien razone con mas lógica pueda superar a quien analice mejor, o que aun quien este mejor inspirado supere al de mas lógica en muchas oportunidades , hace de este juego un tan delicioso como intrascendente arte, y le aseguro perpetua juventud

Cardoso
Posts: 313
Joined: Thu Mar 16, 2006 6:39 pm
Location: Portugal
Full name: Alvaro Cardoso
Contact:

Re: Tactics cannot be very important for chess

Post by Cardoso » Fri Nov 01, 2019 5:49 pm

Dann Corbit wrote:
Fri Nov 01, 2019 7:32 am
It is a cowardly way to play chess. If I think I face an equal, I try to make a tie, and if I think I face a weakling, I do not respect them.
I think that it is better to go for the win at all times, and maintain respect.
So, for instance, if I do not build a wall but leave holes, then we can both go through, and maybe I will lose to a stronger engine.
But then, I really deserve to lose anyway if it is stronger.
And trying to win is the object of the game.
I do not like the contempt approach in either direction. Going for the tie is cowardly. Trying to win only when you think your opponent is a weakling is cowardly. Are we a bunch of cowards playing chess?! Let's put on helmets to pick petunias and pansies.''
Go for it, and if you lose, you probably deserve to lose anyway.
Hi Dann,
I agree with you, in fact sometime ago I posted a reply to a topic in fishcooking, in a way I perhaps shouldn't do, but I was really fed up with all those discussions about contempt:
This is what I said in 1-24-2018
Ivan, you seem to me a very nice guy and all, also you are very prolific in computer chess progarmming. But I'm a bit tired of reading so much about contempt. Wouldn't it be much better spending our time inventing algorithms/heuristics to improve the strength of our engines? I've been reading about contempt "everywhere" as if it is the ultimate heuristic to add some wins to SF or whatever engine. Please, stop this contempt fever and lets use our imagination to invent real algorithms to improve chess engines strength. Not easy? Sure, but wasting our time with contempt won't advance computer chess. Just my two cents.
https://groups.google.com/d/msg/fishcoo ... YGz8QtBQAJ

bob
Posts: 20914
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: Tactics cannot be very important for chess

Post by bob » Fri Nov 01, 2019 5:56 pm

A strong GM, many years ago, wrote something to the effect "Chess is purely about tactics..."

His thinking was the positional play is nothing more than very deep / very long-range tactics. IE what is bad about a backward pawn? It gives you a weakness and your opponent a target. A target that can't move, which will be around for a long time, most likely being lost at some point. In that light, I have to agree with him. At some point in the future we won't need a backward pawn evaluation term because the search will see deeply enough to realize the pawn will be won or lost, period. I have seen a few evaluation terms in my programs over the years that became superfluous.

The ultimate goal is to checkmate the opponent's king. That is purely a tactical goal.

Dann Corbit
Posts: 11169
Joined: Wed Mar 08, 2006 7:57 pm
Location: Redmond, WA USA
Contact:

Re: Tactics cannot be very important for chess

Post by Dann Corbit » Fri Nov 01, 2019 5:58 pm

Cardoso,I think you said it more eloquently than me.
I gave up trying to give suggestions there.
I would point out a crash from bad input, and they would say, "Don't provide bad input."
Well, I process millions upon millions of EPD records and I do not write them myself.
Some .001% or so have errors.
But I have an error checking preprocessor that I use now, so the darn thing won't crash and cost me a machine day worth of analysis
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.

Dann Corbit
Posts: 11169
Joined: Wed Mar 08, 2006 7:57 pm
Location: Redmond, WA USA
Contact:

Re: Tactics cannot be very important for chess

Post by Dann Corbit » Fri Nov 01, 2019 6:03 pm

bob wrote:
Fri Nov 01, 2019 5:56 pm
A strong GM, many years ago, wrote something to the effect "Chess is purely about tactics..."

His thinking was the positional play is nothing more than very deep / very long-range tactics. IE what is bad about a backward pawn? It gives you a weakness and your opponent a target. A target that can't move, which will be around for a long time, most likely being lost at some point. In that light, I have to agree with him. At some point in the future we won't need a backward pawn evaluation term because the search will see deeply enough to realize the pawn will be won or lost, period. I have seen a few evaluation terms in my programs over the years that became superfluous.

The ultimate goal is to checkmate the opponent's king. That is purely a tactical goal.
There is the notion that strategy is just deep tactics.
And as the searches get deeper, the line between tactics and strategy may blur.
But there are some moves that we make which really do not have a tactical goal at the time we make them.
But I guess the bottom line is to always choose the best move(s). Whatever that may be.
Taking ideas is not a vice, it is a virtue. We have another word for this. It is called learning.
But sharing ideas is an even greater virtue. We have another word for this. It is called teaching.

User avatar
mclane
Posts: 17258
Joined: Thu Mar 09, 2006 5:40 pm
Location: US of Europe, germany
Full name: Thorsten Czub
Contact:

Re: Tactics cannot be very important for chess

Post by mclane » Fri Nov 01, 2019 6:05 pm

I want to disagree.

Most moves in a game there is NO tactics.
In those positions you either have a plan, or if you are a computer program, you try to increase your score.

But better then making a more or less random move is, to create a plan. E.g. a plan to mate the opponent king.
Or a plan to bring the isolani to a promotion. Or a plan to storm the king side.

So. A plan can be tactical motive. But there is no need for it.
A plan can have all kind of topics.

Important is: in a position where no tactical shot can be made is a plan a move that creates a longer idea,
Often you have to sacrifice something to get the idea through,

Not always sacking a piece. Often a tempo, or you move a developed piece backwards.
Or you make the 2nd best move or the third best move.

All to bring the plan forward.


IMO chess Programs on the Moment play chess by solving a position. And if they cannot find a key move, they make a move that increases
The positional values. But this is IMO far from playing chess.
They were able to reach 3000 elo with this technic, But do they play chess ? No.

Playing chess is to have a plan.

It’s unimportant if the plan works.
Important is to have one.
What seems like a fairy tale today may be reality tomorrow.
Here we have a fairy tale of the day after tomorrow....

bob
Posts: 20914
Joined: Mon Feb 27, 2006 6:30 pm
Location: Birmingham, AL

Re: Tactics cannot be very important for chess

Post by bob » Fri Nov 01, 2019 6:45 pm

But what is the "plan"? Has to be to checkmate the opponent's king. If you can't see a way to force that, then you seek a way to reach a position where that goal is more reachable, even if you can't see it. But think about it, the game of chess is not infinite. So it is theoretically possible to search from beginning to end. At that point there would be no planning, no positional evaluation, just a sequence of moves that lead you to a won position. I'd have to call that a "tactical outcome"

Today, you are completely correct, until we get to positions that are forced wins (EGTB and such). But one day...

User avatar
mclane
Posts: 17258
Joined: Thu Mar 09, 2006 5:40 pm
Location: US of Europe, germany
Full name: Thorsten Czub
Contact:

Re: Tactics cannot be very important for chess

Post by mclane » Fri Nov 01, 2019 6:57 pm

Right, the major plan is to mate.


Of course this goal is often far. Far away.
That can be seen by Stockfish and other clones doing draw over draw over draw.
If the goal would be easily seen, they would see it in their 30+ search depth, or ?
But they don’t see it.


Now the question is: what to do.

The programs behave like David levy.

Do nothing but do it well.


But is this chess ?


No.

CHESS is to TRY a plan.

I want to see chess programs trying plans.

Show me one.
That is NOT relying on search depth but instead on a plan.
Like a human Beeing.

We should setup a prize for the first chess program that is NOT seeing anything in tree bit instead analysing position and creating a plan and trying it. It’s unimportant if the plan works. That comes later. But it should try it.
What seems like a fairy tale today may be reality tomorrow.
Here we have a fairy tale of the day after tomorrow....

jp
Posts: 1345
Joined: Mon Apr 23, 2018 5:54 am

Re: Tactics cannot be very important for chess

Post by jp » Fri Nov 01, 2019 7:01 pm

bob wrote:
Fri Nov 01, 2019 5:56 pm
A strong GM, many years ago, wrote something to the effect "Chess is purely about tactics..."
And a weak GM quoted and corrected his statement to "Chess is 99% calculation...", which is even more true for computer programs. There is no concept of "positional" versus "tactical" for engines. Those are human concepts. For engines, it's just calculation.

User avatar
mclane
Posts: 17258
Joined: Thu Mar 09, 2006 5:40 pm
Location: US of Europe, germany
Full name: Thorsten Czub
Contact:

Re: Tactics cannot be very important for chess

Post by mclane » Fri Nov 01, 2019 7:06 pm

Nonsense

If you have no plan you can play Backgammon

What do you think why the draw rate of computerchess games is increasing significantly
What seems like a fairy tale today may be reality tomorrow.
Here we have a fairy tale of the day after tomorrow....

Post Reply